Optimization of MR Fingerprinting (MRF) for Quantitative MRI

定量 MRI 的 MR 指纹 (MRF) 优化

• 批准号: 8696434
• 负责人: Mark Griswold
• 金额: $ 50.98万
• 依托单位: CASE WESTERN RESERVE UNIVERSITY
• 依托单位国家: 美国
• 财政年份: 2014
• 资助国家: 美国
• 起止时间: 2014-03-15 至 2018-02-28
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/8696434
• 关键词: Algorithms; Biological Neural Networks; Characteristics; Clinical; Clinical Medicine; Data; Data Set; Development; Diagnosis; Diagnostic; Diffusion; Disease; Fingerprint; Fingers; Generations; Goals; Health; Healthcare; Human; Image; Imagery; Individual; Link; Magnetic Resonance; Magnetic Resonance Imaging; Maps; Medical; Medical Imaging; Methods; Monitor; Motion; Normal tissue morphology; Patients; Pattern Recognition; Physiologic pulse; Physiology; Printing; Process; Property; Randomized; Reliance; Research; Running; Scanning; Series; Signal Transduction; Specificity; System; Technology; Time; Tissues; Translating; base; disease diagnosis; experience; flexibility; imaging modality; improved; interest; magnetic field; non-invasive imaging; novel strategies; public health relevance; reconstruction; research study; signal processing; soft tissue; tool

DESCRIPTION (provided by applicant): MRI has become a key component of clinical medicine because it offers exquisite images of soft tissues that can be made sensitive to almost any disease process. However, it is increasingly clear that the current paradigm of MRI has run up against significant limitations that are, in many cases, unavoidable. Current clinical MRI is effective but practically restricted to a qualitative depiction of a limited set of tissue propertis which are visualized through a series of different acquisitions. Further, the reliance on human interpretation limits the total amount of information that can be assessed in any given exam. Here we propose a radically different paradigm that could dramatically increase the efficiency and specificity of MRI by taking a completely different approach to image acquisition, post-processing and visualization. This new class of methods, Magnetic Resonance Fingerprinting (MRF), overcomes traditional limitations by fully embracing the concept of signal incoherence at the core of compressed sensing. Because of the richness of the parameters that can be analyzed, MR- Fingerprinting methods could augment current interpretation to improve diagnosis and monitoring of disease. Indeed, we are proposing an altogether new approach to medical imaging that directly collects anatomically informed quantitative information about disease state and physiology.

描述(申请人提供)：MRI已成为临床医学的一个关键组成部分，因为它提供了可以对几乎任何疾病过程敏感的软组织的精美图像。然而，越来越清楚的是，目前的核磁共振成像范式遇到了重大限制，在许多情况下，这些限制是不可避免的。目前的临床MRI是有效的，但实际上局限于对一组有限的组织属性的定性描述，这些属性通过一系列不同的采集进行可视化。此外，对人工翻译的依赖限制了在任何给定考试中可以评估的总信息量。在这里，我们提出了一个完全不同的范式，可以通过采取完全不同的方法来获取图像、后处理和可视化，从而显著提高MRI的效率和特异性。这类新的方法，磁共振指纹识别(MRF)，通过完全接受压缩感知核心的信号不一致性的概念，克服了传统的限制。由于可以分析的参数的丰富性，磁共振指纹方法可以增强当前的解释，以改善对疾病的诊断和监测。事实上，我们正在提出一种全新的医学成像方法，直接收集有关疾病状态和生理的解剖学信息。